Objectives: Histological assessment of patients with chronic hepatitis C infection is no longer performed routinely; consequently, a simple test is needed to identify patients with significant hepatic fibrosis.
Methods: Data were collected, retrospectively, on 923 consecutive patients undergoing percutaneous liver biopsy for chronic hepatitis C at King's College Hospital between 1 January 2000 and 30 June 2006; 602 patients were accepted to form the training set and a further 105 patients to form the validation set.
Results: On liver biopsy, 132 (22%) had cirrhosis (Ishak F5-6) in the training set and 19 (18%) in the validation set. Factors found by multivariate analysis to be associated with fibrosis in the training set were used to construct the King's Score: age x aspartate aminotransferase x international normalized ratio / platelets. Area under receiver operating characteristic curves for predicting cirrhosis and significant fibrosis (F3-6) were 0.91 and 0.79, respectively. A King's Score of greater than or equal to 16.7 predicted cirrhosis in 34% of patients (odds ratio 36.2, 95% confidence interval, 22.0-59.6; P<0.0001) with sensitivity 86%, specificity 80% and a high negative predictive value of 96%; a score greater than or equal to 12.3 predicted F3-6 (odds ratio 33.9, 95% confidence interval, 15.2-34.4; P<0.001). The validation set confirmed the utility of this index, area under receiver operating characteristic curves 0.94 and 0.89 for cirrhosis and F3-6, respectively.
Conclusion: The King's Score is a simple and accurate index for predicting cirrhosis in chronic hepatitis C. Patients with a score of less than 16.7 have a low risk of cirrhosis.